From Man to Machine: Federal Aviation Administration Slashing Human Airport Weather Observers; Trusts Weather Observing Machines
Weather is one of those jobs that much of society doesn’t seem to care about, until they care about it. The science is continuously watered down, mocked, and hyped for personal and financial interest by large broadcasting companies, phone application creators, and folks who simply don’t think you need to be a scientist anymore because we have the MyRadar smartphone app and the all-knowing YouTube.
That’s all fine & dandy…until someone gets hurt. And at airports, it’s usually more than just one person if the weather acts up and there’s no weatherman there to alert pilots & air traffic controllers of weather dangers…which can sometimes be invisible.
The Federal Aviation Administration (FAA) has taken aim at weather observers at the majority of towered airfields (generally commercial-airlines airports) and ruled their jobs either not important enough, or too expensive. They’ve suggested two fixes. First, train some already task-saturated, stressed-out air traffic controllers to observe the weather (while controlling airplanes, too!), or simply just let the weather observing machine do the work.
Let me battle both of this unacceptable & terrifying options, one at time. I’ve worked side-by-side with Air Traffic Controllers. I’ve observed weather professionally. I’ve even maintained & fixed weather observing machines (from here on, known as Automated Surface Observing System, or ASOS), and boy they need a lot of work.
First, let me explain the job of an Air Traffic Controller. One of the most stressful jobs on the planet, controllers are given strict medical standards and work/rest cycles. At many commercial airports, especially in metropolitan areas, controllers speak to hundreds of airlines and private pilots during a normal shift, and route them around each other, out of restricted flight zones, and somehow manage to get them all on the ground…or in the air safely…sometimes simultaneously with other aircraft in the case of an airport with multiple runways. During peak airline rush hours, this is bumper-to-bumper traffic (with strict interval spacing required between aircraft) in three dimensions, with no traffic signals or braking capabilities. And that’s on a perfect weather day. I’ll let you try to imagine a stormy…or foggy...or icy day, and put yourself in the shoes of an air traffic controller.
But, the FAA must think that it’s not that difficult…controllers are already juggling thousands of lives on airplanes crammed to the minute by tight schedules in congested airspace, why not walk outside, lick their finger, and stick it in the air and guess the wind direction while they’re at it?
Now, these weather observation machines…ASOS…are great for some weather things…like temperature, humidity, precipitation, and pressure. They can also be pretty decent with wind speed & direction, and I rarely ever find issues with these aspects, unless the computer program it transmits data with crashes, or there is an electrical issue (which can usually be resolved with a simple reset).
Some of the biggest issues I deal with on nearly a daily basis from ASOS systems are related to cloud coverage, height, surface horizontal visibility, and lightning detection. Typically speaking, other than turbulence, these weather features that the ASOS struggles with are the most impactful to aviation (plus wind speed/direction). So, it's important that we get them right, even if the machine doesn't.
The FAA has suggested that we just let the ASOS do its job. Great idea, on a good weather day. Bad idea on a day that we are expecting clouds, fog, or thunderstorms.
Here’s the limitations to each of these weather sensing components on the ASOS, and I’ll let you be the judge on whether you, as an airline passenger, would feel if your life was left in the hands of a machine, rather than a meteorologist as you are trying to fly through bad weather.
Cloud height is extremely important to aviation for a number of reasons. Higher clouds can cause icing concerns for your aircraft. If de-icing equipment isn’t working, it could freeze the aircraft controls in place. High clouds also can show turbulence, at times. Turbulence is invisible (it’s literally just wind differences), but the effects of it can be seen in clouds, if we're lucky.
Low clouds generally pose more of a threat to aviation, largely because they restrict a pilot’s view of the ground. When you can’t see the ground…you don’t always know how close you are to it. When you can’t see a mountain ahead…that could be dangerous. When you can’t see other airplanes…bad. And when you can’t see the runway because the clouds are low…how are you supposed to land?
Cloud coverage can easily be determined by a weather observer. They can see an unobstructed view of the celestial dome…and report any clouds they see...low, mid, high clouds…what types of clouds they are, and so forth. Clouds levels are important, because they dictate the status of the airfield. Ever see the rotating white and green beacon at an airport? It’s on at night so pilots can see it in the dark, but it stays on during the day if the clouds and visibility are poor.
On the ASOS, the ceilometer cannot “see” the entire celestial dome, but rather only what is right above it. One end of the runway be socked in the fog, but the ceilometer may not even know. It only send a laser beam straight up and waits for a response, taking a 10-minute average of responses to capture cloud coverage and cloud height. It can’t see very high, either…often capping out at about 12,000 feet. One of the biggest problems I see with these instruments, outside of horrible coverage (since it only looks straight up), is that the lenses are easily scratched and the entire instrument won’t report any clouds, even if it’s completely overcast. And, even if it can see the clouds, it has no way of telling what kind of cloud it is…a thunderstorm cloud (big deal), a cirrus cloud (not a big deal), or even a wave cloud (turbulence- big deal).
Visibility, like cloud heights, is very important. If a pilot can’t see, he or she can’t fly…at least without the help of on-board cockpit instruments. The visibility sensor on the ASOS somehow was calibrated to understand visibility out to 10 miles, but only with two opposing lenses sitting less than two feet apart from each other. Not only do these lenses collect dust (and falsely report low visibility on clear days), but they only tell pilots what the visibility really is in the two feet between the two lenses, not across the runway or the flight approach to the airport. And, I’ve picked more than one spider web off of these too, which causes erroneous data.
Lightning detection. Thunderstorms are bad, especially if you’re flying. Benjamin Franklin proved this by flying a key on a kite, but apparently the FAA is having trouble understanding the gravity of the importance of thunderstorm observing. Although airplanes can sustain a lightning strike in many cases, pilots tend to avoid thunderstorms in an effort to mitigate the risk for any electronic or structural failure, and because thunderstorms can produce extremely strong turbulence, icing, rain, hail…just everything you don’t want to fly through. Lightning detectors on the ASOS systems are calibrated to sense electrical charges when a lightning bolt hits the ground…but their sensitivity is marginal at best. And, they tend to not capture lightning in clouds or lightning in the distance. Don’t you think a pilot would want to know that? I would hope my pilot would!
Microbursts & Low Level Wind Shear. These are often invisible, and certainly cannot be observed or sensed by a machine until too late. But, a meteorologist weather observer is trained to identify when one of these very dangerous weather events is occurring or approaching, potentially saving hundreds of lives in the hands of a pilot that was unaware of the threat.
Since the 1960’s, an average of 6% (63) of aircraft crashes have occurred due to weather alone, without any pilot error. An average of 58% of crashes were attributed to pilot error, in which one would at least entertain that perhaps poor weather may have played a factor in pilot judgment.
Aside from the safety of aviation, surface weather observations, regardless of whether they are disseminated by a meteorologist or a machine, are used to initiate computer forecast models used for global forecasting, including those for severe weather such as hurricanes, tornadoes, and blizzards. Additionally, they are recorded in permanent record for use in climatology- the long-term study of climates and statistical weather analysis.
The FAA argues the price for human weather observers…which I’ve read is somewhere between $20M and $60M annually, in total. In my estimation, that’s actually a pretty dang good deal to keep planes flying safe. However, the pot of money that this comes from may not necessarily have to come from the federal government…but perhaps the airports themselves. Certainly, they get paid for every airplane that lands on their runway(s). That could afford several weather observers…keep the taxpayer cost down, and avoid aviation weather mishaps due to heavy reliance on extremely flawed automated weather sensor data.
- Meteorologist Dan Schreiber
Who Needs a Meteorologist? I Have A Smartphone! How Flawed Public Trust in Increased Technology Is Literally Destroying Weather Forecasting Jobs
As a meteorologist, I’ve been a long-time opponent to weather cell phone applications. I don’t watch The Weather Channel, I take local news weather reports with a grain of salt, and I purposely still own, and use, a flip-phone.
With high-definition weather info at our fingertips…what’s the point of a weatherman? In fact, many TV stations…even The Weather Company (parent company of The Weather Channel recently bought by IBM) is employing broadcasters, computer technicians, and social scientists as so-called “meteorologists”. Let me be very clear: They are NOT meteorologists. Below is a "meteorologist" opening for a Sinclair Brodcasting TV Station in Pensacola:
But, large companies like IBM are simply allowed to do this because of society’s desire of finger-tip access to data and excitement in a “weather show” over, well, reality.
While computers and technology continue to improve to the point where, on fair-weather days, computers can, within reason, accurately predict the weather without human interaction. I still believe that there is something to be said for a meteorologist’s oversight on even these forecasts, but certainly if there are no clouds in the sky and its July, in the desert, how difficult is it to forecast that afternoon’s temperature? Hot.
But, there are many, many aspects of weather forecasting that computers are far from capable, such as severe weather, like hail storms, flash flooding, and tornadoes. Meteorologists, not computer scientists and TV broadcasters, need to analyze forecast model data and create an actionable forecast for these events.
American society has put too much confidence in technology, and is very quickly neglecting scientific reasoning and understanding. I’m a meteorologist, I examine and analyze the raw, unaltered computer forecasts models every day. It’s just data…a derivative of simply 0’s and 1’s…and sometimes, the output is literally junk, but only a meteorologist would recognize it as junk. If it were ingested into a cell phone application…the output would be wrong, in some cases, dangerously wrong, but the user would have no idea.
Companies like IBM, and nearly every broadcasting agency…are all introducing extremely dangerous solutions. The Weather Company, for instance, hires about four non-meteorologists for every one meteorologist. Even a high-level operations chief at The Weather Company stated in an interview with ArsTechinca.com that “There’s no doubt that the trend in the big picture is less of a pedigree on solely meteorology, and more of a pedigree on meteorology plus analytics or data sciences or other related areas.” In other words, the science doesn’t matter…it’s how many “clicks” we get on our website and how many “likes” we get on Facebook. How much money we bring in. The Weather Company is more interested in how customers use their data than they are about the data itself. Does anyone else see the problem here?
This is scary. I’ll put it in perspective. Could you imagine a tornado warning (data) that was issued not because of a tornado (whether or not it existed), but because a company found it financial conducive? Or better yet, a tornado warning that was not issued (that did occur) because some company had a cell-phone application builder at the desk that doesn’t know a thing about tornadoes?
If we continue to not employ meteorologists in meteorology jobs, this is exactly where we are headed. Big-name weather companies, solely out of financial interest, are attempting to not only overtake the weather forecasting role with technology that falsely eradicates the need of human meteorologists, but also takes a step further in providing decision-making consulting services…all via computer technology.
For instance, if you were a concrete worker, and you wanted to know how strong the wind would blow during the afternoon because you were concerned about concrete drying too rapidly…a phone application would alert you of winds becoming too strong. Great, in theory.
But, there’s two big problems. One, it’s just a computer talking, not an actual meteorologist. The forecast that you would be receiving may not have even been quality-checked by a meteorologist. Second, you can’t discuss your concrete-pouring activities with a smart phone application, but you can with a meteorologist that has a vested interest in you and your business.
The idea that meteorology isn’t very important, and that meteorologists aren’t necessary for weather science and forecasting is absurd. There’s already too stiff of a reliance on technology across-the-board in first-world societies. It is putting folks out of jobs at an exponential rate…mechanics, engineers, store clerks, book & music store employees, retail shop workers…ever wonder why the old blacksmith trade dissolved into thin air? Sure, catch up with the times. But meteorology…a science…will can never be fully comprehended or mastered, only improved upon by skilled meteorologists. It’s not meant to be a TV show or a phone application, for that matter. It’s a serious science that costs thousands of lives annually…and its meteorological skill that is the forefront of communicating deadly weather that should be the prized possession of weather companies, not their fancy web applications.
- Meteorologist Dan Schreiber
Sometimes you don't have to have words to explain how amazing God's creation is. Satellite Images, all from the same time frame on October 24th of a well-developed Low Pressure offshore of the Pacific Northwest Coast. Enjoy.
-Meteorologist Dan Schreiber
Some days, most days, I shake my head a bit when I see the news or read the headlines about the utter lack of scholastic intelligence that society expects out of high school and, heaven forbid, college graduates. Of course, this isn’t any breaking news…but with media and technology these days, it’s just more apparent.
The United States, once prized for its educational system, is slowly falling behind other nations both in Asia and Europe in smarts. Why is there such a strong surge of technological advances going on in Japan and Korea? Enormous and successful companies like Samsung, Toyota, LG, Hyundai, Toshiba…I could keep going. My last post even discusses atmospheric forecasting computer models that are far superior to American models being created in Europe.
This shouldn’t be surprising at all, really. Why have nations like Japan and Korea, Norway, Sweden…why have they stolen the technological and scientific fame that America once held as it’s prized possession? Perhaps it has to do with scholastic environment and academic standards.
In America, between the “No Child Left Behind” Act and the introduction of the hideous, disgraceful Common Core State-Standards…teachers are literally forced to assign passing grades to failing students. Not only that, but certain adopted curriculum standards simply outlaw free-thought…taking after the famed, ruthless, industrialist John D. Rockefeller’s quote, “I don't want a nation of thinkers. I want a nation of workers.” Rockefeller didn’t say that because it was good for the nation (he wasn’t a politician), but rather it was good for himself (maybe should have been a politician?).
As the United States is falling behind academically to other nations, our academic leaders are introducing standards that simply do not allow the formation of scientific hypothesis, theories, and eventually, facts, which don’t align with these standards…like Portland, Oregon’s ban on “textbooks and material casting doubt on human effects on climate change”. We’re talking about science! But, apparently Portland Public Schools already thinks that there is nothing left to discover in science.
One of my favorite movies of all time is Apollo 13. It’s a great movie because it shows what America was capable of, even before all of the technology we have these days. America didn’t put a man on the moon because some rocket scientist was gently passed through grade school because “no child left behind”. In the movie, Houston Space Center was able to return three astronauts from the dark side of the moon back to Earth with virtually no power, no oxygen, no computer, and outer space temperature extremes. Those space scientists weren’t high-school flunkies given a free pass to scoot them along their way.
Meanwhile, these days, the United States scoffs at North Korea for failing rocket launches, but our own missile defense system which would be used to protect an attack, according to the LA Times, has proven unreliable…and we have our own SpaceX rocket blowing up before we even press the launch button. When we do get a spacecraft launched…in the case of 1999’s Mars Climate Orbiter, someone input the wrong units into the engine thrusters meant to put the craft in stable orbit around the planet, and led to the craft’s burning up in Mars’ upper-atmosphere. This is a mistake that could have been caught by a high-schooler (one that legitimately passes middle-school), but wasn’t caught by American rocket scientists.
Too make things even a little dicier, some candidates for President of the United States in 2016 have mentioned “free college”, and that everyone should be given a chance to succeed. While I agree that opportunities to succeed are great, and that America should be the place to succeed, if we are, in the words of The Washington Post, “giving [students] a diploma, no matter how fraudulent, [because it] might provide them with a chance to get some kind of job and, eventually, as they mature, sort themselves out”, we’re in a heap of trouble in this nation. Folks who aren't even supposed to be graduating high school will be graduating college and entering a workforce, disguised as "educated", and we expect to see success in America?
My case has been validated by the New York Post, “As the American economy sputters along and many people live paycheck-to-paycheck, economists say a highly-skilled workforce is key to economic recovery.”
So, I have a couple of questions for our politicians & society:
- Meteorologist Dan Schreiber
The New York Times interviewed Professor Cliff Mass, a renowned meteorologist employed by the University of Washington- Seatlle, and he believes that many, if not all of these inaccuracies can be attributed to the mis-managed, underfunded, and behind-the-times technology of the National Oceanic & Atmospheric Administration (NOAA), the parent organization which the National Weather Service (NWS) belongs.
Robby Herman, a blogger from BigThink.com, who according to his short bio, knows little about any science, not to mention dynamic weather modelling, also blogged about the aforementioned New York Times article on Professor Mass, re-stating the idea that if only NOAA were better funded and more technologically advanced, like several European counterparts, America's weather forecasts would improve to the level of certain European nations weather forecasts, like England.
Mr. Herman's blog post has many inaccuracies, and I can attest to that as an operational meteorologist. But, he does make a point that there are several foreign weather forecast models that have the physics package capable of producing highly accurate forecasts, above and beyond what many domestic American forecast models are capable of on a routine basis.
One of these is produced by the European Center for Medium-Range Forecasts (ECMWF), and it is indeed a fine forecast model. What neither Professor Mass nor Mr. Herman stated was that many operational meteorologists, including those with the National Weather Service, use this foreign forecast model side-by-side with the Canadian Meteorological Center (CMC) forecast model. and with domestic American forecast models, like the similar (but lower resolution) Global Forecast System (GFS) model. So, while some foreign forecast models may outperform, at times, American models...I hardly believe that we can simply blame lack of technology for incorrect weather forecasts.
Should our weather forecast computer models be improved to match the quality of other nations? Yes, and they are. NOAA is well on it's way to introducing a new weather forecasting model, known as the FV3, that should replace the current GFS by 2020 (but keep the same name, apparently).
Meanwhile, the United States Air Force has purchased access to the Unified Model, also known as the UKMET, from the English as well, instead of contributing to the improvement of in-house American forecast modelling systems. In fact, the Air Force has even begun to remove domestic weather models from it's forecasting user-interface...in essence an "out of site, out of mind" approach. The domestic models they are abandoning? One of them is the North American Model (NAM), a domestic NOAA-produced forecasting model with high-resolution capability- the same capabilities that Mr. Herman's blog stated were advantageous to the European forecasting models which are outperforming our lower-resolution GFS model. So, there's obviously more to the story than just simply computing power, perhaps having something to do with Meteosat-7, a European satellite nearing the end of it's life cycle over the war-torn Middle East, which the Department of Defense has no reasonable plans to replace. .
Computing power is certainly a contributing factor to both good and bad weather forecasts. Meteorologists simply just can't do their job without computers, and the better the technology, the higher the potential for an accurate weather forecast. But, can we just blame the computer?
Absolutely not! Even Professor Mass admitted that during a far-over-predicted snow storm in 2015, the National Weather Service forecaster "broke all the rules I teach my students". So, the computing power and technology was there - even Professor Mass, who is a top-advocate for improved forecasting models and higher-tech computing power, admits that the forecasters were to blame for inaccuracy. Computers, government funding, and technology are all simple scapegoat answers for the underlying problem.
This is the problem with technology. While it's beautiful, it also creates complacency & laziness.
Case in point, look at smart phones these days. How much information is stored on that device that used to be stored in people's heads? Notes, phone numbers, directions...all replaced by technology. I've met some folks who didn't even know their spouse's phone number by memory, but it was stored on their phone. In other words, we have so [too] much of a reliance on technology that we're losing track of our education, training, and personnel capabilities.
Inherently, the real problem with forecasting weather is the lack of responsibility in forecasting agencies, the lack of passion within many forecasters, and the lack of accountability by weather forecasting authorities. Weather forecasting is a battle, and even the best get forecasts wrong, at times. Unfortunately, and I've personally witnessed this through my career, some meteorologists have the mindset that it's "ok to be wrong"...as if weather forecasting is a baseball game and you get paid well to hit the ball and actually make it to first base during only 30% of your at-bats.
I'll put it this way. If you were in a doctor's office because of chest pain, and the doctor diagnosed you with a bruised rib, but the real problem was a heart malfunction, what would you blame, the medical technology or the doctor? The doctor, of course! He or she should have known that there was a substantial difference between the symptoms or a bruised rib and that of a heart attack, without even consulting medical computer technology.
The same can be said about meteorology and weather forecasting. How can we blame computing power for failing to forecast weather correctly? Simply by looking at current weather information (i.e. the symptoms), any decent weather forecaster has the training to assess the capabilities, reliability, and biases of the weather forecast models to be used to create a forecast.
The problem is the lack of internal accountability. Are weather forecasters being responsible and using their skills and training? Do the weather forecasts make sense, from a scientific perspective? How many trained eyes are there quality-checking forecasts before they are made official? What are some common mistakes made, and what can be done to eliminate them? Good questions to ask...and I can personally say that this sort of thing rarely happens.
Just like doctors have a responsibility to their patients, meteorologists must be held at the same level of responsibility to their clients...and the general public.
-Meteorologist Dan Schreiber
Dan Schreiber is a freelance meteorologist with experience